Tag-based annotation creates better avatars
This work addresses the problem of efficient and accurate avatar creation for users and developers, offering an incremental improvement over existing rendering systems by reducing annotation noise and adaptation costs.
The paper tackles the challenge of creating avatars from human images by addressing the difficulty users face in tuning numerous parameters and the high label noise in training data due to annotator struggles. It proposes a tag-based annotation method that improves annotator agreement, leads to more consistent machine learning predictions, and reduces the cost of adding new rendering systems.
Avatar creation from human images allows users to customize their digital figures in different styles. Existing rendering systems like Bitmoji, MetaHuman, and Google Cartoonset provide expressive rendering systems that serve as excellent design tools for users. However, twenty-plus parameters, some including hundreds of options, must be tuned to achieve ideal results. Thus it is challenging for users to create the perfect avatar. A machine learning model could be trained to predict avatars from images, however the annotators who label pairwise training data have the same difficulty as users, causing high label noise. In addition, each new rendering system or version update requires thousands of new training pairs. In this paper, we propose a Tag-based annotation method for avatar creation. Compared to direct annotation of labels, the proposed method: produces higher annotator agreements, causes machine learning to generates more consistent predictions, and only requires a marginal cost to add new rendering systems.